This was an open prospective one-group single-center 12-month study in patients with HTN, who attended HTN Center of Excellence in the city of St. Petersburg, Russian Federation. Of those who met inclusion/exclusion criteria were consecutively enrolled in a study with BPTM.
Inclusion criteria: uncontrolled HTN (office SBP level ≥ 140 mmHg or self-reported home SBP ≥ 135 mm Hg and ongoing treatment with at least one antihypertensive drug in the last month). Exclusion criteria were as follows: age above 80 years, symptomatic cardiovascular or other major comorbidities requiring close medical monitoring (<3-month periods), pregnancy, significant cognitive impairment, and active or acute mental problem. All the patients were to have a smartphone/tablet with instant high-speed Internet access (Wi-Fi or cellular 3 or 4 G).
The study design as follows: for the first 3 months patients are being actively monitored and counseled. The rest 9 months of the study represent passive follow-up which implies no mandatory office visits, and patients might continue BPTM at their sole intention. The final visit should have been conducted at 12-month point. For this analysis we used the data of the first 3-month active period.
All patients signed the informed consent document at the baseline visit. The study was carried according to the ICH GCP standards and the Helsinki Declaration of the World Medical Association. The study protocol was approved by the local ethics committee.
Office blood pressure
Office BP measurements were performed at baseline and at 3-month visits according to the ESC/ESH Guidelines . Each time, BP was measured with automatic oscillometric device Omron M3 Expert (HEM 7132-ALRU, Kyoto, Japan) after 3-5 minutes of quiet rest in sitting position with the back and (dominant) arm supported. An appropriate bladder cuff was used, encircling at least 80% of the arm. Three serial BP readings were taken 1–2 minutes apart, and the average of the last two readings was displayed. We used two variations of SBP control in the office: a cut-off of <140 mmHg or TR within 120-139 mmHg.
Blood pressure telemonitoring and online counseling
A free simple website and a mobile application was used for patient–physician communication as well as storage and exchange of medical information. Detailed information on the hybrid telehealth solution can be found elsewhere . In brief, patients used the mobile application, while web-based software was installed on office computers at the clinical site. Each patient was managed by the same physician throughout the follow-up period. At baseline, patients were registered in the program and their accounts were linked to doctors’ ones. The interface allowed the patient to manually input the HBP data. A text chat window was available for remote consultations (an unlimited number with a 24- to 72-hour timeframe for a physician to reply).
Home blood pressure monitoring
Initially it was recommended to record BP by a validated device in the morning and evening, 3 times in a row with 1-minute intervals, before meals and drug intake, for 3 to 7 days. Patients were advised to manually enter the last two BP readings into the electronic HBP diary. HBPM tutorials and the list of validated devices (STRIDE-BP ) were available for patients in the dedicated ‘Support’ section of the BPTM app. The supervising physician was advised to check HBP readings after the first week and then to adjust HBPM accordingly (using the text chat). Results of HBPM guided antihypertensive treatment titration. We used two variations of SBP control at home: a cut-off of <135 mmHg or TR within 110-130 mmHg.
Proportion of home systolic blood pressure readings in target range
Here we introduce the measure called proportion of home SBP readings in target range (sBPiTR). For this analysis we express this variable as the percent of home SBPs which fall within 110 and 130 mmHg in a certain time window (Equation 1)
Due to an observational nature of the study and our previous experience with BPTM, home monitoring might be chaotic. So, we did not expect patients would monitor BP on the daily basis but rather skip some days (or even weeks) taking breaks.
Hence, we applied 2 scenario towards calculation of sBPiTR: (1) to analyze all available BP measurements in the electronic diary except the very first day or (2) to analyze BP readings discarding every first day if HBPM was interrupted ≥7 days.
We then divided sBPiTR (0-3 months) into 4 groups (quartiles): high rate of home SBP control (75-100%), more than half of SBP readings in TR (50–74%), less than half of SBP readings in TR (25–49%), and low rate of home SBP control (0-24%).
Descriptive statistics included median and interquartile range, IQR for continuous variables (the data were non-normally distributed). We applied a frequency analysis (the χ2 test) to assess the contingency between counts and proportions. We applied MacNemar’s test for the paired nominal variables. Continuous variables were compared by Mann-Whitney U test, and Wilcoxon rank-sum test was used for paired parameters. Weighted Cohen’s kappa coefficient was used as a measure of inter-rater reliability between office/home SBP (nominal variable) and different sBPiTR quartiles. Kappa coefficients were interpreted according to McHugh . Spearman’s Rho (rs) coefficient was used to assess the association between the variables.
Multivariate logistic regression analysis was used to assess the associations between controlled HTN (per office (1) or home (2) SBP) as dependent categorical variables (in TR/not in TR) and sBPiTR (main independent categorical variable), with the adjustment for age, sex, number of antihypertensive drugs, baseline office SBP (included as covariates). Only the best-case sBPiTR scenario was taken as a potential predictor and the results are presented as the odds ratio (OR) and 95% confidence interval (95% CI).
Two-sided p values <0.05 were considered signiﬁcant.
Statistical analyses were carried out by two authors (MI, ME) using SPSS version 23 (IBM SPSS, Chicago, IL, United States), Python Software Foundation (Python Language Reference, version 2.7; available at http://www.python.org) and jamovi. (the jamovi project, version 1.6 for MacOS) retrieved from https://www.jamovi.org.